line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
1
|
|
|
|
|
|
|
package Lingua::Identify; |
2
|
|
|
|
|
|
|
|
3
|
7
|
|
|
7
|
|
168609
|
use 5.006; |
|
7
|
|
|
|
|
23
|
|
|
7
|
|
|
|
|
260
|
|
4
|
7
|
|
|
7
|
|
38
|
use strict; |
|
7
|
|
|
|
|
12
|
|
|
7
|
|
|
|
|
210
|
|
5
|
7
|
|
|
7
|
|
38
|
use warnings; |
|
7
|
|
|
|
|
16
|
|
|
7
|
|
|
|
|
193
|
|
6
|
|
|
|
|
|
|
|
7
|
7
|
|
|
7
|
|
4928
|
use utf8; |
|
7
|
|
|
|
|
49
|
|
|
7
|
|
|
|
|
39
|
|
8
|
7
|
|
|
7
|
|
241
|
use base 'Exporter'; |
|
7
|
|
|
|
|
14
|
|
|
7
|
|
|
|
|
3187
|
|
9
|
|
|
|
|
|
|
|
10
|
|
|
|
|
|
|
our %EXPORT_TAGS = |
11
|
|
|
|
|
|
|
( |
12
|
|
|
|
|
|
|
all => [ qw( |
13
|
|
|
|
|
|
|
langof |
14
|
|
|
|
|
|
|
langof_file |
15
|
|
|
|
|
|
|
confidence |
16
|
|
|
|
|
|
|
get_all_methods |
17
|
|
|
|
|
|
|
activate_all_languages |
18
|
|
|
|
|
|
|
deactivate_all_languages |
19
|
|
|
|
|
|
|
get_all_languages |
20
|
|
|
|
|
|
|
get_active_languages |
21
|
|
|
|
|
|
|
get_inactive_languages |
22
|
|
|
|
|
|
|
is_active |
23
|
|
|
|
|
|
|
is_valid_language |
24
|
|
|
|
|
|
|
activate_language |
25
|
|
|
|
|
|
|
deactivate_language |
26
|
|
|
|
|
|
|
set_active_languages |
27
|
|
|
|
|
|
|
name_of |
28
|
|
|
|
|
|
|
) |
29
|
|
|
|
|
|
|
], |
30
|
|
|
|
|
|
|
language_identification => [ qw( |
31
|
|
|
|
|
|
|
langof |
32
|
|
|
|
|
|
|
langof_file |
33
|
|
|
|
|
|
|
confidence |
34
|
|
|
|
|
|
|
get_all_methods |
35
|
|
|
|
|
|
|
) |
36
|
|
|
|
|
|
|
], |
37
|
|
|
|
|
|
|
|
38
|
|
|
|
|
|
|
language_manipulation => [ qw( |
39
|
|
|
|
|
|
|
activate_all_languages |
40
|
|
|
|
|
|
|
deactivate_all_languages |
41
|
|
|
|
|
|
|
get_all_languages |
42
|
|
|
|
|
|
|
get_active_languages |
43
|
|
|
|
|
|
|
get_inactive_languages |
44
|
|
|
|
|
|
|
is_active |
45
|
|
|
|
|
|
|
is_valid_language |
46
|
|
|
|
|
|
|
activate_language |
47
|
|
|
|
|
|
|
deactivate_language |
48
|
|
|
|
|
|
|
set_active_languages |
49
|
|
|
|
|
|
|
name_of |
50
|
|
|
|
|
|
|
) |
51
|
|
|
|
|
|
|
], |
52
|
|
|
|
|
|
|
); |
53
|
|
|
|
|
|
|
|
54
|
|
|
|
|
|
|
our @EXPORT_OK = ( @{ $EXPORT_TAGS{'all'} } ); |
55
|
|
|
|
|
|
|
our @EXPORT = qw(); |
56
|
|
|
|
|
|
|
|
57
|
|
|
|
|
|
|
our $VERSION = '0.57_1'; |
58
|
|
|
|
|
|
|
|
59
|
|
|
|
|
|
|
|
60
|
|
|
|
|
|
|
# DEFAULT VALUES # |
61
|
|
|
|
|
|
|
our %default_methods = qw/smallwords 2 prefixes3 1 suffixes3 1 ngrams3 1 letters 0.5/; |
62
|
|
|
|
|
|
|
my $default_maxsize = 1_000_000; |
63
|
|
|
|
|
|
|
my %default_extractfrom = qw/head 1/; |
64
|
|
|
|
|
|
|
|
65
|
|
|
|
|
|
|
=encoding utf8 |
66
|
|
|
|
|
|
|
|
67
|
|
|
|
|
|
|
=head1 NAME |
68
|
|
|
|
|
|
|
|
69
|
|
|
|
|
|
|
Lingua::Identify - Language identification |
70
|
|
|
|
|
|
|
|
71
|
|
|
|
|
|
|
=head1 SYNOPSIS |
72
|
|
|
|
|
|
|
|
73
|
|
|
|
|
|
|
use Lingua::Identify qw(:language_identification); |
74
|
|
|
|
|
|
|
$a = langof($textstring); # gives the most probable language |
75
|
|
|
|
|
|
|
|
76
|
|
|
|
|
|
|
or the complete way: |
77
|
|
|
|
|
|
|
|
78
|
|
|
|
|
|
|
@a = langof($textstring); # gives pairs of languages / probabilities |
79
|
|
|
|
|
|
|
# sorted from most to least probable |
80
|
|
|
|
|
|
|
|
81
|
|
|
|
|
|
|
%a = langof($textstring); # gives a hash of language / probability |
82
|
|
|
|
|
|
|
|
83
|
|
|
|
|
|
|
or the expert way (see section OPTIONS, under HOW TO PERFORM IDENTIFICATION) |
84
|
|
|
|
|
|
|
|
85
|
|
|
|
|
|
|
$a = langof( { method => [qw/smallwords prefix2 suffix2/] }, $text); |
86
|
|
|
|
|
|
|
|
87
|
|
|
|
|
|
|
$a = langof( { 'max-size' => 3_000_000 }, $text); |
88
|
|
|
|
|
|
|
|
89
|
|
|
|
|
|
|
$a = langof( { 'extract_from' => ( 'head' => 1, 'tail' => 2)}, $text); |
90
|
|
|
|
|
|
|
|
91
|
|
|
|
|
|
|
=head1 DESCRIPTION |
92
|
|
|
|
|
|
|
|
93
|
|
|
|
|
|
|
B |
94
|
|
|
|
|
|
|
|
95
|
|
|
|
|
|
|
C identifies the language a given string or file is |
96
|
|
|
|
|
|
|
written in. |
97
|
|
|
|
|
|
|
|
98
|
|
|
|
|
|
|
See section WHY LINGUA::IDENTIFY for a list of C's strong |
99
|
|
|
|
|
|
|
points. |
100
|
|
|
|
|
|
|
|
101
|
|
|
|
|
|
|
See section KNOWN LANGUAGES for a list of available languages and HOW TO |
102
|
|
|
|
|
|
|
PERFORM IDENTIFICATION to know how to really use this module. |
103
|
|
|
|
|
|
|
|
104
|
|
|
|
|
|
|
If you're in a hurry, jump to section EXAMPLES, way down below. |
105
|
|
|
|
|
|
|
|
106
|
|
|
|
|
|
|
Also, don't forget to read the following section, IMPORTANT WARNING. |
107
|
|
|
|
|
|
|
|
108
|
|
|
|
|
|
|
=head1 A WARNING ON THE ACCURACY OF LANGUAGE IDENTIFICATION METHODS |
109
|
|
|
|
|
|
|
|
110
|
|
|
|
|
|
|
Take a word that exists in two different languages, take a good look at it and |
111
|
|
|
|
|
|
|
answer this question: "What language does this word belong to?". |
112
|
|
|
|
|
|
|
|
113
|
|
|
|
|
|
|
You can't give an answer like "Language X", right? You can only say it looks |
114
|
|
|
|
|
|
|
like any of a set of languages. |
115
|
|
|
|
|
|
|
|
116
|
|
|
|
|
|
|
Similarly, it isn't always easy to identify the language of a text if the only |
117
|
|
|
|
|
|
|
two active languages are very similar. |
118
|
|
|
|
|
|
|
|
119
|
|
|
|
|
|
|
Now that we've taken out of the way the warning that language identification |
120
|
|
|
|
|
|
|
is not 100% accurate, please keep reading the documentation. |
121
|
|
|
|
|
|
|
|
122
|
|
|
|
|
|
|
=head1 WHY LINGUA::IDENTIFY |
123
|
|
|
|
|
|
|
|
124
|
|
|
|
|
|
|
You might be wondering why you should use Lingua::Identify instead of any other |
125
|
|
|
|
|
|
|
tool for language identification. |
126
|
|
|
|
|
|
|
|
127
|
|
|
|
|
|
|
Here's a list of Lingua::Identify's strong points: |
128
|
|
|
|
|
|
|
|
129
|
|
|
|
|
|
|
=over 6 |
130
|
|
|
|
|
|
|
|
131
|
|
|
|
|
|
|
=item * it's free and it's open-source; |
132
|
|
|
|
|
|
|
|
133
|
|
|
|
|
|
|
=item * it's portable (it's Perl, which means it will work in lots of different |
134
|
|
|
|
|
|
|
platforms); |
135
|
|
|
|
|
|
|
|
136
|
|
|
|
|
|
|
=item * unicode support; |
137
|
|
|
|
|
|
|
|
138
|
|
|
|
|
|
|
=item * 4 different methods of language identification and growing (see |
139
|
|
|
|
|
|
|
METHODS OF LANGUAGE IDENTIFICATION for more details on this one); |
140
|
|
|
|
|
|
|
|
141
|
|
|
|
|
|
|
=item * it's a module, which means you can easily write your own application |
142
|
|
|
|
|
|
|
(be it CGI, TK, whatever) around it; |
143
|
|
|
|
|
|
|
|
144
|
|
|
|
|
|
|
=item * it comes with I, which means you don't actually need to |
145
|
|
|
|
|
|
|
write your own application around it; |
146
|
|
|
|
|
|
|
|
147
|
|
|
|
|
|
|
=item * it's flexible (at the moment, you can actually choose the |
148
|
|
|
|
|
|
|
methods to use and their relevance, the max size of input to analyze |
149
|
|
|
|
|
|
|
each time and which part(s) of the input to analyze) |
150
|
|
|
|
|
|
|
|
151
|
|
|
|
|
|
|
=item * it supports big inputs (through the 'max-size' and |
152
|
|
|
|
|
|
|
'extract_from' options) |
153
|
|
|
|
|
|
|
|
154
|
|
|
|
|
|
|
=item * it's easy to deal with languages (you can activate and |
155
|
|
|
|
|
|
|
deactivate the ones you choose whenever you want to, which can improve |
156
|
|
|
|
|
|
|
your times and accuracy); |
157
|
|
|
|
|
|
|
|
158
|
|
|
|
|
|
|
=item * it's maintained. |
159
|
|
|
|
|
|
|
|
160
|
|
|
|
|
|
|
=back |
161
|
|
|
|
|
|
|
|
162
|
|
|
|
|
|
|
=cut |
163
|
|
|
|
|
|
|
|
164
|
|
|
|
|
|
|
# initialization |
165
|
|
|
|
|
|
|
|
166
|
|
|
|
|
|
|
our (@all_languages,@active_languages,%languages,%regexen,@methods); |
167
|
|
|
|
|
|
|
BEGIN { |
168
|
|
|
|
|
|
|
|
169
|
7
|
|
|
7
|
|
6391
|
use Class::Factory::Util; |
|
7
|
|
|
|
|
4797
|
|
|
7
|
|
|
|
|
42
|
|
170
|
7
|
|
|
7
|
|
1992
|
for ( Lingua::Identify->subclasses() ) { |
171
|
196
|
100
|
|
|
|
17969
|
/^[A-Z]{2}(?:_[A-Z]{2})?$/ || next; |
172
|
189
|
|
|
|
|
13787
|
eval "require Lingua::Identify::$_ ;"; |
173
|
189
|
50
|
|
|
|
1678
|
if ($languages{_versions}{lc $_} <= 0.02) { |
174
|
0
|
|
|
|
|
0
|
for my $k (keys %languages) { |
175
|
0
|
0
|
|
|
|
0
|
delete($languages{$k}{lc $_}) if exists $languages{$k}{lc $_}; |
176
|
|
|
|
|
|
|
} |
177
|
|
|
|
|
|
|
} |
178
|
|
|
|
|
|
|
} |
179
|
|
|
|
|
|
|
|
180
|
7
|
|
|
|
|
94
|
@all_languages = @active_languages = keys %{$languages{_names}}; |
|
7
|
|
|
|
|
267
|
|
181
|
|
|
|
|
|
|
|
182
|
7
|
|
|
|
|
14057
|
@methods = qw/smallwords/; |
183
|
|
|
|
|
|
|
|
184
|
|
|
|
|
|
|
} |
185
|
|
|
|
|
|
|
|
186
|
|
|
|
|
|
|
=head1 HOW TO PERFORM IDENTIFICATION |
187
|
|
|
|
|
|
|
|
188
|
|
|
|
|
|
|
=head2 langof |
189
|
|
|
|
|
|
|
|
190
|
|
|
|
|
|
|
To identify the language a given text is written in, use the I function. |
191
|
|
|
|
|
|
|
To get a single value, do: |
192
|
|
|
|
|
|
|
|
193
|
|
|
|
|
|
|
$language = langof($text); |
194
|
|
|
|
|
|
|
|
195
|
|
|
|
|
|
|
To get the most probable language and also the percentage of its probability, |
196
|
|
|
|
|
|
|
do: |
197
|
|
|
|
|
|
|
|
198
|
|
|
|
|
|
|
($language, $probability) = langof($text); |
199
|
|
|
|
|
|
|
|
200
|
|
|
|
|
|
|
If you want a hash where each active language is mapped into its percentage, |
201
|
|
|
|
|
|
|
use this: |
202
|
|
|
|
|
|
|
|
203
|
|
|
|
|
|
|
%languages = langof($text); |
204
|
|
|
|
|
|
|
|
205
|
|
|
|
|
|
|
=cut |
206
|
|
|
|
|
|
|
|
207
|
|
|
|
|
|
|
sub langof { |
208
|
72
|
|
|
72
|
1
|
17213
|
my %config = (); |
209
|
72
|
100
|
|
|
|
338
|
%config = (%config, %{+shift}) if ref($_[0]) eq 'HASH'; |
|
42
|
|
|
|
|
133
|
|
210
|
|
|
|
|
|
|
|
211
|
|
|
|
|
|
|
=head3 OPTIONS |
212
|
|
|
|
|
|
|
|
213
|
|
|
|
|
|
|
I can also be given some configuration parameters, in this way: |
214
|
|
|
|
|
|
|
|
215
|
|
|
|
|
|
|
$language = langof(\%config, $text); |
216
|
|
|
|
|
|
|
|
217
|
|
|
|
|
|
|
These parameters are detailed here: |
218
|
|
|
|
|
|
|
|
219
|
|
|
|
|
|
|
=over 6 |
220
|
|
|
|
|
|
|
|
221
|
|
|
|
|
|
|
=item * B |
222
|
|
|
|
|
|
|
|
223
|
|
|
|
|
|
|
When the size of the input exceeds the C'max-size', C analyzes |
224
|
|
|
|
|
|
|
only the beginning of the file. You can specify which part of the file |
225
|
|
|
|
|
|
|
is analyzed with the 'extract-from' option: |
226
|
|
|
|
|
|
|
|
227
|
|
|
|
|
|
|
langof( { 'extract_from' => 'tail' } , $text ); |
228
|
|
|
|
|
|
|
|
229
|
|
|
|
|
|
|
Possible values are 'head' and 'tail' (for now). |
230
|
|
|
|
|
|
|
|
231
|
|
|
|
|
|
|
You can also specify more than one part of the file, so that text is |
232
|
|
|
|
|
|
|
extracted from those parts: |
233
|
|
|
|
|
|
|
|
234
|
|
|
|
|
|
|
langof( { 'extract_from' => [ 'head', 'tail' ] } , $text ); |
235
|
|
|
|
|
|
|
|
236
|
|
|
|
|
|
|
(this will be useful when more than two possibilities exist) |
237
|
|
|
|
|
|
|
|
238
|
|
|
|
|
|
|
You can also specify different values for each part of the file (not |
239
|
|
|
|
|
|
|
necessarily for all of them: |
240
|
|
|
|
|
|
|
|
241
|
|
|
|
|
|
|
langof( { 'extract_from' => { head => 40, tail => 60 } } , $text); |
242
|
|
|
|
|
|
|
|
243
|
|
|
|
|
|
|
The line above, for instance, retrives 40% of the text from the |
244
|
|
|
|
|
|
|
beginning and 60% from the end. Note, however, that those values are |
245
|
|
|
|
|
|
|
not percentages. You'd get the same behavior with: |
246
|
|
|
|
|
|
|
|
247
|
|
|
|
|
|
|
langof( { 'extract_from' => { head => 80, tail => 120 } } , $text); |
248
|
|
|
|
|
|
|
|
249
|
|
|
|
|
|
|
The percentages would be the same. |
250
|
|
|
|
|
|
|
|
251
|
|
|
|
|
|
|
=item * B |
252
|
|
|
|
|
|
|
|
253
|
|
|
|
|
|
|
By default, C analyzes only 1,000,000 bytes. You can specify |
254
|
|
|
|
|
|
|
how many bytes (at the most) can be analyzed (if not enough exist, the |
255
|
|
|
|
|
|
|
whole input is still analyzed). |
256
|
|
|
|
|
|
|
|
257
|
|
|
|
|
|
|
langof( { 'max-size' => 2000 }, $text); |
258
|
|
|
|
|
|
|
|
259
|
|
|
|
|
|
|
If you want all the text to be analyzed, set max-size to 0: |
260
|
|
|
|
|
|
|
|
261
|
|
|
|
|
|
|
langof( { 'max-size' => 0 }, $text); |
262
|
|
|
|
|
|
|
|
263
|
|
|
|
|
|
|
See also C. |
264
|
|
|
|
|
|
|
|
265
|
|
|
|
|
|
|
=item * B |
266
|
|
|
|
|
|
|
|
267
|
|
|
|
|
|
|
You can choose which method or methods to use, and also the relevance of each of |
268
|
|
|
|
|
|
|
them. |
269
|
|
|
|
|
|
|
|
270
|
|
|
|
|
|
|
To choose a single method to use: |
271
|
|
|
|
|
|
|
|
272
|
|
|
|
|
|
|
langof( {method => 'smallwords' }, $text); |
273
|
|
|
|
|
|
|
|
274
|
|
|
|
|
|
|
To choose several methods: |
275
|
|
|
|
|
|
|
|
276
|
|
|
|
|
|
|
langof( {method => [qw/prefixes2 suffixes2/]}, $text); |
277
|
|
|
|
|
|
|
|
278
|
|
|
|
|
|
|
To choose several methods and give them different weight: |
279
|
|
|
|
|
|
|
|
280
|
|
|
|
|
|
|
langof( {method => {smallwords => 0.5, ngrams3 => 1.5} }, $text); |
281
|
|
|
|
|
|
|
|
282
|
|
|
|
|
|
|
To see the list of available methods, see section METHODS OF LANGUAGE |
283
|
|
|
|
|
|
|
IDENTIFICATION. |
284
|
|
|
|
|
|
|
|
285
|
|
|
|
|
|
|
If no method is specified, the configuration for this parameter is the |
286
|
|
|
|
|
|
|
following (this might change in the future): |
287
|
|
|
|
|
|
|
|
288
|
|
|
|
|
|
|
method => { |
289
|
|
|
|
|
|
|
smallwords => 0.5, |
290
|
|
|
|
|
|
|
prefixes2 => 1, |
291
|
|
|
|
|
|
|
suffixes3 => 1, |
292
|
|
|
|
|
|
|
ngrams3 => 1.3 |
293
|
|
|
|
|
|
|
}; |
294
|
|
|
|
|
|
|
|
295
|
|
|
|
|
|
|
=item * B |
296
|
|
|
|
|
|
|
|
297
|
|
|
|
|
|
|
By default, C assumes C mode, but others are |
298
|
|
|
|
|
|
|
available. |
299
|
|
|
|
|
|
|
|
300
|
|
|
|
|
|
|
In C mode, instead of actually calculating anything, |
301
|
|
|
|
|
|
|
C only does the preparation it has to and then |
302
|
|
|
|
|
|
|
returns a bunch of information, including the list of the active |
303
|
|
|
|
|
|
|
languages, the selected methods, etc. It also returns the text meant |
304
|
|
|
|
|
|
|
to be analised. |
305
|
|
|
|
|
|
|
|
306
|
|
|
|
|
|
|
Do be warned that, with I, the dummy mode still reads the |
307
|
|
|
|
|
|
|
files, it simply doesn't calculate language. |
308
|
|
|
|
|
|
|
|
309
|
|
|
|
|
|
|
langof( { 'mode' => 'dummy' }, $text); |
310
|
|
|
|
|
|
|
|
311
|
|
|
|
|
|
|
This returns something like this: |
312
|
|
|
|
|
|
|
|
313
|
|
|
|
|
|
|
{ 'methods' => { 'smallwords' => '0.5', |
314
|
|
|
|
|
|
|
'prefixes2' => '1', |
315
|
|
|
|
|
|
|
}, |
316
|
|
|
|
|
|
|
'config' => { 'mode' => 'dummy' }, |
317
|
|
|
|
|
|
|
'max-size' => 1000000, |
318
|
|
|
|
|
|
|
'active-languages' => [ 'es', 'pt' ], |
319
|
|
|
|
|
|
|
'text' => $text, |
320
|
|
|
|
|
|
|
'mode' => 'dummy', |
321
|
|
|
|
|
|
|
} |
322
|
|
|
|
|
|
|
|
323
|
|
|
|
|
|
|
=back |
324
|
|
|
|
|
|
|
|
325
|
|
|
|
|
|
|
=cut |
326
|
|
|
|
|
|
|
|
327
|
|
|
|
|
|
|
# select the methods |
328
|
72
|
100
|
|
|
|
682
|
my %methods = defined $config{'method'} ? _make_hash($config{'method'}) |
329
|
|
|
|
|
|
|
: %default_methods; |
330
|
|
|
|
|
|
|
|
331
|
|
|
|
|
|
|
# select max-size |
332
|
72
|
100
|
|
|
|
290
|
my $maxsize = defined $config{'max-size'} ? $config{'max-size'} |
333
|
|
|
|
|
|
|
: $default_maxsize; |
334
|
|
|
|
|
|
|
|
335
|
|
|
|
|
|
|
# get the text |
336
|
72
|
|
|
|
|
452
|
my $text = join "\n", @_; |
337
|
72
|
100
|
|
|
|
208
|
return wantarray ? () : undef unless $text; |
|
|
100
|
|
|
|
|
|
338
|
|
|
|
|
|
|
|
339
|
|
|
|
|
|
|
# this is the support for big files; if the input is bigger than the $maxsize, we act |
340
|
70
|
100
|
100
|
|
|
986
|
if ($maxsize < length $text && $maxsize != 0) { |
341
|
|
|
|
|
|
|
# select extract_from |
342
|
1
|
50
|
|
|
|
9
|
my %extractfrom = defined $config{'extract_from'} ? _make_hash($config{'extract_from'}) |
343
|
|
|
|
|
|
|
: %default_extractfrom; |
344
|
1
|
|
|
|
|
4
|
my $total_weight = 0; |
345
|
1
|
|
|
|
|
5
|
for (keys %extractfrom) { |
346
|
1
|
50
|
33
|
|
|
8
|
if ($_ eq 'head' or $_ eq 'tail') { |
347
|
1
|
|
|
|
|
5
|
$total_weight += $extractfrom{$_}; |
348
|
1
|
|
|
|
|
3
|
next; |
349
|
|
|
|
|
|
|
} |
350
|
|
|
|
|
|
|
else { |
351
|
0
|
|
|
|
|
0
|
delete $extractfrom{$_}; |
352
|
|
|
|
|
|
|
} |
353
|
|
|
|
|
|
|
} |
354
|
1
|
|
|
|
|
11
|
for (keys %extractfrom) { |
355
|
1
|
|
|
|
|
7
|
$extractfrom{$_} = $extractfrom{$_} / $total_weight; |
356
|
|
|
|
|
|
|
} |
357
|
|
|
|
|
|
|
|
358
|
1
|
|
50
|
|
|
6
|
$extractfrom{'head'} ||= 0; |
359
|
1
|
|
50
|
|
|
8
|
$extractfrom{'tail'} ||= 0; |
360
|
|
|
|
|
|
|
|
361
|
1
|
|
|
|
|
5
|
my $head = int $maxsize * $extractfrom{'head'}; |
362
|
1
|
|
|
|
|
4
|
my $tail = length($text) - $head - int $maxsize * $extractfrom{'tail'}; |
363
|
1
|
|
|
|
|
7
|
substr( $text, $head, $tail, ''); |
364
|
|
|
|
|
|
|
} |
365
|
|
|
|
|
|
|
|
366
|
|
|
|
|
|
|
# dummy mode exits here |
367
|
70
|
|
100
|
|
|
430
|
$config{'mode'} ||= 'normal'; |
368
|
70
|
100
|
|
|
|
219
|
if ($config{'mode'} eq 'dummy') { |
369
|
|
|
|
|
|
|
return { |
370
|
7
|
|
|
|
|
25
|
'method' => \%methods, |
371
|
|
|
|
|
|
|
'max-size' => $maxsize, |
372
|
|
|
|
|
|
|
'config' => \%config, |
373
|
|
|
|
|
|
|
'active-languages' => [ sort (get_active_languages()) ], |
374
|
|
|
|
|
|
|
'text' => $text, |
375
|
|
|
|
|
|
|
'mode' => $config{'mode'}, |
376
|
|
|
|
|
|
|
}; |
377
|
|
|
|
|
|
|
} |
378
|
|
|
|
|
|
|
|
379
|
|
|
|
|
|
|
# use the methods |
380
|
63
|
|
|
|
|
75
|
my (%result, $total); |
381
|
63
|
|
|
|
|
188
|
for (keys %methods) { |
382
|
295
|
|
|
|
|
449
|
my %temp_result; |
383
|
|
|
|
|
|
|
|
384
|
295
|
100
|
|
|
|
2891
|
if (/^letters$/) { |
|
|
100
|
|
|
|
|
|
|
|
100
|
|
|
|
|
|
|
|
100
|
|
|
|
|
|
|
|
50
|
|
|
|
|
|
385
|
58
|
|
|
|
|
213
|
%temp_result = _langof_by_used_letters('letters', $text); |
386
|
|
|
|
|
|
|
} |
387
|
|
|
|
|
|
|
elsif (/^smallwords$/) { |
388
|
60
|
|
|
|
|
294
|
%temp_result = _langof_by_word_method('smallwords', $text); |
389
|
|
|
|
|
|
|
} |
390
|
|
|
|
|
|
|
elsif (/^(prefixes[1-4])$/) { |
391
|
58
|
|
|
|
|
249
|
%temp_result = _langof_by_prefix_method($1, $text); |
392
|
|
|
|
|
|
|
} |
393
|
|
|
|
|
|
|
elsif (/^(suffixes[1-4])$/) { |
394
|
61
|
|
|
|
|
277
|
%temp_result = _langof_by_suffix_method($1, $text); |
395
|
|
|
|
|
|
|
} |
396
|
|
|
|
|
|
|
elsif (/^(ngrams[1-4])$/) { |
397
|
58
|
|
|
|
|
250
|
%temp_result = _langof_by_ngram_method($1, $text); |
398
|
|
|
|
|
|
|
} |
399
|
|
|
|
|
|
|
|
400
|
295
|
|
|
|
|
3729
|
for my $l (keys %temp_result) { |
401
|
4733
|
|
|
|
|
7726
|
my $temp = $temp_result{$l} * $methods{$_}; |
402
|
4733
|
|
|
|
|
5684
|
$result{$l} += $temp; |
403
|
4733
|
|
|
|
|
7054
|
$total += $temp; |
404
|
|
|
|
|
|
|
} |
405
|
|
|
|
|
|
|
} |
406
|
|
|
|
|
|
|
|
407
|
|
|
|
|
|
|
# report the results |
408
|
1587
|
50
|
|
|
|
3516
|
my @result = ( |
409
|
5600
|
|
|
|
|
6829
|
map { ( $_, ($total ? $result{$_} / $total : 0)) } |
410
|
63
|
|
|
|
|
434
|
sort { $result{$b} <=> $result{$a} } keys %result |
411
|
|
|
|
|
|
|
); |
412
|
|
|
|
|
|
|
|
413
|
63
|
100
|
|
|
|
1705
|
return wantarray ? @result : $result[0]; |
414
|
|
|
|
|
|
|
} |
415
|
|
|
|
|
|
|
|
416
|
|
|
|
|
|
|
sub _make_hash { |
417
|
11
|
|
|
11
|
|
17
|
my %hash; |
418
|
11
|
|
|
|
|
20
|
my $temp = shift; |
419
|
11
|
|
|
|
|
29
|
for (ref($temp)) { |
420
|
11
|
100
|
|
|
|
51
|
if (/^HASH$/) { |
|
|
100
|
|
|
|
|
|
421
|
1
|
|
|
|
|
2
|
%hash = %{$temp}; |
|
1
|
|
|
|
|
7
|
|
422
|
|
|
|
|
|
|
} |
423
|
|
|
|
|
|
|
elsif (/^ARRAY$/) { |
424
|
1
|
|
|
|
|
2
|
for (@{$temp}) { |
|
1
|
|
|
|
|
3
|
|
425
|
2
|
|
|
|
|
7
|
$hash{$_}++; |
426
|
|
|
|
|
|
|
} |
427
|
|
|
|
|
|
|
} |
428
|
|
|
|
|
|
|
else { |
429
|
9
|
|
|
|
|
43
|
$hash{$temp} = 1; |
430
|
|
|
|
|
|
|
} |
431
|
|
|
|
|
|
|
} |
432
|
11
|
|
|
|
|
51
|
%hash; |
433
|
|
|
|
|
|
|
} |
434
|
|
|
|
|
|
|
|
435
|
|
|
|
|
|
|
=head2 langof_file |
436
|
|
|
|
|
|
|
|
437
|
|
|
|
|
|
|
I works just like I, with the exception that it |
438
|
|
|
|
|
|
|
reveives filenames instead of text. It reads these texts (if existing |
439
|
|
|
|
|
|
|
and readable, of course) and parses its content. |
440
|
|
|
|
|
|
|
|
441
|
|
|
|
|
|
|
Currently, I assumes the files are regular text. This may |
442
|
|
|
|
|
|
|
change in the future and the files might be scanned to check their |
443
|
|
|
|
|
|
|
filetype and then parsed to extract only their textual content (which |
444
|
|
|
|
|
|
|
should be pretty useful so that you can perform language |
445
|
|
|
|
|
|
|
identification, say, in HTML files, or PDFs). |
446
|
|
|
|
|
|
|
|
447
|
|
|
|
|
|
|
To identify the language a file is written in: |
448
|
|
|
|
|
|
|
|
449
|
|
|
|
|
|
|
$language = langof_file($path); |
450
|
|
|
|
|
|
|
|
451
|
|
|
|
|
|
|
To get the most probable language and also the percentage of its probability, |
452
|
|
|
|
|
|
|
do: |
453
|
|
|
|
|
|
|
|
454
|
|
|
|
|
|
|
($language, $probability) = langof_file($path); |
455
|
|
|
|
|
|
|
|
456
|
|
|
|
|
|
|
If you want a hash where each active language is mapped into its percentage, |
457
|
|
|
|
|
|
|
use this: |
458
|
|
|
|
|
|
|
|
459
|
|
|
|
|
|
|
%languages = langof_file($path); |
460
|
|
|
|
|
|
|
|
461
|
|
|
|
|
|
|
If you pass more than one file to I, they will all be |
462
|
|
|
|
|
|
|
read and their content merged and then parsed for language |
463
|
|
|
|
|
|
|
identification. |
464
|
|
|
|
|
|
|
|
465
|
|
|
|
|
|
|
=cut |
466
|
|
|
|
|
|
|
|
467
|
|
|
|
|
|
|
sub langof_file { |
468
|
31
|
|
|
31
|
1
|
21768
|
my %config = (); |
469
|
31
|
100
|
|
|
|
156
|
if (ref($_[0]) eq 'HASH') {%config = (%config, %{+shift})} |
|
2
|
|
|
|
|
5
|
|
|
2
|
|
|
|
|
11
|
|
470
|
|
|
|
|
|
|
|
471
|
|
|
|
|
|
|
=head3 OPTIONS |
472
|
|
|
|
|
|
|
|
473
|
|
|
|
|
|
|
I accepts all the options I does, so refer to |
474
|
|
|
|
|
|
|
those first (up in this document). |
475
|
|
|
|
|
|
|
|
476
|
|
|
|
|
|
|
$language = langof_file(\%config, $path); |
477
|
|
|
|
|
|
|
|
478
|
|
|
|
|
|
|
I currently only reads the first 10,000 bytes of each |
479
|
|
|
|
|
|
|
file. |
480
|
|
|
|
|
|
|
|
481
|
|
|
|
|
|
|
You can force an input encoding with C<< { encoding => 'ISO-8859-1' } >> |
482
|
|
|
|
|
|
|
in the configuration hash. |
483
|
|
|
|
|
|
|
|
484
|
|
|
|
|
|
|
=cut |
485
|
|
|
|
|
|
|
|
486
|
|
|
|
|
|
|
# select max-size |
487
|
31
|
50
|
|
|
|
142
|
my $maxsize = defined $config{'max-size'} ? $config{'max-size'} |
488
|
|
|
|
|
|
|
: $default_maxsize; |
489
|
|
|
|
|
|
|
|
490
|
31
|
|
|
|
|
97
|
my @files = @_; |
491
|
31
|
|
|
|
|
74
|
my $text = ''; |
492
|
|
|
|
|
|
|
|
493
|
31
|
|
|
|
|
72
|
for my $file (@files) { |
494
|
|
|
|
|
|
|
#-r and -e or next; |
495
|
32
|
100
|
|
|
|
114
|
if (exists($config{encoding})) { |
496
|
1
|
50
|
|
1
|
|
45
|
open(FILE, "<:encoding($config{encoding})", $file) or next; |
|
1
|
|
|
|
|
9
|
|
|
1
|
|
|
|
|
3
|
|
|
1
|
|
|
|
|
8
|
|
497
|
|
|
|
|
|
|
} else { |
498
|
31
|
50
|
|
|
|
1739
|
open(FILE, "<:utf8", $file) or next; |
499
|
|
|
|
|
|
|
} |
500
|
32
|
|
|
|
|
13707
|
local $/ = \$maxsize; |
501
|
32
|
|
|
|
|
1980
|
$text .= ; |
502
|
32
|
|
|
|
|
812
|
close(FILE); |
503
|
|
|
|
|
|
|
} |
504
|
|
|
|
|
|
|
|
505
|
31
|
|
|
|
|
134
|
return langof(\%config,$text); |
506
|
|
|
|
|
|
|
} |
507
|
|
|
|
|
|
|
|
508
|
|
|
|
|
|
|
=head2 confidence |
509
|
|
|
|
|
|
|
|
510
|
|
|
|
|
|
|
After getting the results into an array, its first element is the most probable |
511
|
|
|
|
|
|
|
language. That doesn't mean it is very probable or not. |
512
|
|
|
|
|
|
|
|
513
|
|
|
|
|
|
|
You can find more about the likeliness of the results to be accurate by |
514
|
|
|
|
|
|
|
computing its confidence level. |
515
|
|
|
|
|
|
|
|
516
|
|
|
|
|
|
|
use Lingua::Identify qw/:language_identification/; |
517
|
|
|
|
|
|
|
my @results = langof($text); |
518
|
|
|
|
|
|
|
my $confidence_level = confidence(@results); |
519
|
|
|
|
|
|
|
# $confidence_level now holds a value between 0.5 and 1; the higher that |
520
|
|
|
|
|
|
|
# value, the more accurate the results seem to be |
521
|
|
|
|
|
|
|
|
522
|
|
|
|
|
|
|
The formula used is pretty simple: p1 / (p1 + p2) , where p1 is the |
523
|
|
|
|
|
|
|
probability of the most likely language and p2 is the probability of |
524
|
|
|
|
|
|
|
the language which came in second. A couple of examples to illustrate |
525
|
|
|
|
|
|
|
this: |
526
|
|
|
|
|
|
|
|
527
|
|
|
|
|
|
|
English 50% Portuguese 10% ... |
528
|
|
|
|
|
|
|
|
529
|
|
|
|
|
|
|
confidence level: 50 / (50 + 10) = 0.83 |
530
|
|
|
|
|
|
|
|
531
|
|
|
|
|
|
|
Another example: |
532
|
|
|
|
|
|
|
|
533
|
|
|
|
|
|
|
Spanish 30% Portuguese 10% ... |
534
|
|
|
|
|
|
|
|
535
|
|
|
|
|
|
|
confidence level: 30 / (25 + 30) = 0.55 |
536
|
|
|
|
|
|
|
|
537
|
|
|
|
|
|
|
French 10% German 5% ... |
538
|
|
|
|
|
|
|
|
539
|
|
|
|
|
|
|
confidence level: 10 / (10 + 5) = 0.67 |
540
|
|
|
|
|
|
|
|
541
|
|
|
|
|
|
|
As you can see, the first example is probably the most accurate one. |
542
|
|
|
|
|
|
|
Are there any doubts? The English language has five times the |
543
|
|
|
|
|
|
|
probability of the second language. |
544
|
|
|
|
|
|
|
|
545
|
|
|
|
|
|
|
The second example is a bit more tricky. 55% confidence. The |
546
|
|
|
|
|
|
|
confidence level is always above 50%, for obvious reasons. 55% doesn't |
547
|
|
|
|
|
|
|
make anyone confident in the results, and one shouldn't be, with |
548
|
|
|
|
|
|
|
results such as these. |
549
|
|
|
|
|
|
|
|
550
|
|
|
|
|
|
|
Notice the third example. The confidence level goes up to 67%, but the |
551
|
|
|
|
|
|
|
probability of French is of mere 10%. So what? It's twice as much as |
552
|
|
|
|
|
|
|
the second language. The low probability may well be caused by a great |
553
|
|
|
|
|
|
|
number of languages in play. |
554
|
|
|
|
|
|
|
|
555
|
|
|
|
|
|
|
=cut |
556
|
|
|
|
|
|
|
|
557
|
|
|
|
|
|
|
sub confidence { |
558
|
62
|
100
|
66
|
62
|
1
|
90196
|
defined $_[1] and $_[1] or return 0; |
559
|
59
|
100
|
33
|
|
|
427
|
defined $_[3] and $_[3] or return 1; |
560
|
58
|
|
|
|
|
359
|
$_[1] / ($_[1] + $_[3]); |
561
|
|
|
|
|
|
|
} |
562
|
|
|
|
|
|
|
|
563
|
|
|
|
|
|
|
=head2 get_all_methods |
564
|
|
|
|
|
|
|
|
565
|
|
|
|
|
|
|
Returns a list comprised of all the available methods for language |
566
|
|
|
|
|
|
|
identification. |
567
|
|
|
|
|
|
|
|
568
|
|
|
|
|
|
|
=cut |
569
|
|
|
|
|
|
|
|
570
|
|
|
|
|
|
|
sub get_all_methods { |
571
|
1
|
|
|
1
|
1
|
28
|
qw/smallwords |
572
|
|
|
|
|
|
|
prefixes1 prefixes2 prefixes3 prefixes4 |
573
|
|
|
|
|
|
|
suffixes1 suffixes2 suffixes3 suffixes4 |
574
|
|
|
|
|
|
|
ngrams1 ngrams2 ngrams3 ngrams4/ |
575
|
|
|
|
|
|
|
} |
576
|
|
|
|
|
|
|
|
577
|
|
|
|
|
|
|
=head1 LANGUAGE IDENTIFICATION IN GENERAL |
578
|
|
|
|
|
|
|
|
579
|
|
|
|
|
|
|
Language identification is based in patterns. |
580
|
|
|
|
|
|
|
|
581
|
|
|
|
|
|
|
In order to identify the language a given text is written in, we repeat a given |
582
|
|
|
|
|
|
|
process for each active language (see section LANGUAGES MANIPULATION); in that |
583
|
|
|
|
|
|
|
process, we look for common patterns of that language. Those patterns can be |
584
|
|
|
|
|
|
|
prefixes, suffixes, common words, ngrams or even sequences of words. |
585
|
|
|
|
|
|
|
|
586
|
|
|
|
|
|
|
After repeating the process for each language, the total score for each of them |
587
|
|
|
|
|
|
|
is then used to compute the probability (in percentage) for each language to be |
588
|
|
|
|
|
|
|
the one of that text. |
589
|
|
|
|
|
|
|
|
590
|
|
|
|
|
|
|
=cut |
591
|
|
|
|
|
|
|
|
592
|
|
|
|
|
|
|
sub _langof_by_method { |
593
|
237
|
|
|
237
|
|
900
|
my ($method, $elements, $text) = @_; |
594
|
237
|
|
|
|
|
341
|
my (%result, $total); |
595
|
|
|
|
|
|
|
|
596
|
237
|
|
|
|
|
703
|
for my $language (get_active_languages()) { |
597
|
6346
|
|
|
|
|
6932
|
for (keys %{$languages{$method}{$language}}) { |
|
6346
|
|
|
|
|
52534
|
|
598
|
173900
|
100
|
|
|
|
321310
|
if (exists $$elements{$_}) { |
599
|
19274
|
|
|
|
|
50134
|
$result{$language} += |
600
|
|
|
|
|
|
|
$$elements{$_} * ${languages{$method}{$language}{$_}}; |
601
|
19274
|
|
|
|
|
43806
|
$total += |
602
|
|
|
|
|
|
|
$$elements{$_} * ${languages{$method}{$language}{$_}}; |
603
|
|
|
|
|
|
|
} |
604
|
|
|
|
|
|
|
} |
605
|
|
|
|
|
|
|
} |
606
|
|
|
|
|
|
|
|
607
|
3167
|
50
|
|
|
|
7917
|
my @result = ( |
608
|
9268
|
|
|
|
|
11927
|
map { ( $_, ($total ? $result{$_} / $total : 0)) } |
609
|
237
|
|
|
|
|
2047
|
sort { $result{$b} <=> $result{$a} } keys %result |
610
|
|
|
|
|
|
|
); |
611
|
|
|
|
|
|
|
|
612
|
237
|
50
|
|
|
|
11136
|
return wantarray ? @result : $result[0]; |
613
|
|
|
|
|
|
|
} |
614
|
|
|
|
|
|
|
|
615
|
|
|
|
|
|
|
=head1 METHODS OF LANGUAGE IDENTIFICATION |
616
|
|
|
|
|
|
|
|
617
|
|
|
|
|
|
|
C currently comprises four different ways for language |
618
|
|
|
|
|
|
|
identification, in a total of thirteen variations of those. |
619
|
|
|
|
|
|
|
|
620
|
|
|
|
|
|
|
The available methods are the following: B, B, |
621
|
|
|
|
|
|
|
B, B, B, B, B, |
622
|
|
|
|
|
|
|
B, B, B, B, B and B. |
623
|
|
|
|
|
|
|
|
624
|
|
|
|
|
|
|
Here's a more detailed explanation of each of those ways and those methods |
625
|
|
|
|
|
|
|
|
626
|
|
|
|
|
|
|
=head2 Small Word Technique - B |
627
|
|
|
|
|
|
|
|
628
|
|
|
|
|
|
|
The "Small Word Technique" searches the text for the most common words of each |
629
|
|
|
|
|
|
|
active language. These words are usually articles, pronouns, etc, which happen |
630
|
|
|
|
|
|
|
to be (usually) the shortest words of the language; hence, the method name. |
631
|
|
|
|
|
|
|
|
632
|
|
|
|
|
|
|
This is usually a good method for big texts, especially if you happen to have |
633
|
|
|
|
|
|
|
few languages active. |
634
|
|
|
|
|
|
|
|
635
|
|
|
|
|
|
|
=cut |
636
|
|
|
|
|
|
|
|
637
|
|
|
|
|
|
|
sub _langof_by_word_method { |
638
|
60
|
|
|
60
|
|
291
|
my ($method, $text) = @_; |
639
|
|
|
|
|
|
|
|
640
|
|
|
|
|
|
|
sub _words_count { |
641
|
60
|
|
|
60
|
|
152
|
my ($words, $text) = @_; |
642
|
60
|
|
|
|
|
20186
|
for my $word (split /[\s\n]+/, $text) { |
643
|
17632
|
|
|
|
|
33905
|
$words->{$word}++ |
644
|
|
|
|
|
|
|
} |
645
|
|
|
|
|
|
|
} |
646
|
|
|
|
|
|
|
|
647
|
60
|
|
|
|
|
154
|
my %words; |
648
|
60
|
|
|
|
|
262
|
_words_count(\%words, $text); |
649
|
60
|
|
|
|
|
1593
|
return _langof_by_method($method, \%words, $text); |
650
|
|
|
|
|
|
|
} |
651
|
|
|
|
|
|
|
|
652
|
|
|
|
|
|
|
=head2 Used Letters - B |
653
|
|
|
|
|
|
|
|
654
|
|
|
|
|
|
|
In this technique used letters are compared with the letters from each |
655
|
|
|
|
|
|
|
language. This method is not meant to be used alone, but to help |
656
|
|
|
|
|
|
|
other methods to differentiate between dialects. |
657
|
|
|
|
|
|
|
|
658
|
|
|
|
|
|
|
=cut |
659
|
|
|
|
|
|
|
|
660
|
|
|
|
|
|
|
sub _langof_by_used_letters { |
661
|
58
|
|
|
58
|
|
259
|
my ($method, $text) = @_; |
662
|
|
|
|
|
|
|
|
663
|
58
|
|
|
|
|
515
|
my $letters = ngram_counts( { spaces => 0 }, $text, 1); |
664
|
|
|
|
|
|
|
|
665
|
58
|
|
|
|
|
104078
|
my ($total, %result); |
666
|
58
|
|
|
|
|
207
|
for my $language (get_active_languages()) { |
667
|
1566
|
|
|
|
|
2777
|
$result{$language} = 1; |
668
|
1566
|
|
|
|
|
12228
|
for my $l (keys %$letters) { |
669
|
44469
|
100
|
|
|
|
120159
|
if (not exists($languages{letters}{$language}{$l})) { |
670
|
10144
|
|
|
|
|
21640
|
$result{$language} -= .1; |
671
|
|
|
|
|
|
|
} |
672
|
|
|
|
|
|
|
} |
673
|
1566
|
100
|
|
|
|
7144
|
$result{$language} = 0 if $result{$language} < 0; |
674
|
1566
|
|
|
|
|
2517
|
$total += $result{$language}; |
675
|
|
|
|
|
|
|
} |
676
|
1566
|
50
|
|
|
|
3782
|
my @result = ( |
677
|
5069
|
|
|
|
|
6143
|
map { ( $_, ($total ? $result{$_} / $total : 0)) } |
678
|
58
|
|
|
|
|
609
|
sort { $result{$b} <=> $result{$a} } keys %result |
679
|
|
|
|
|
|
|
); |
680
|
|
|
|
|
|
|
|
681
|
58
|
50
|
|
|
|
1678
|
return wantarray ? @result : $result[0]; |
682
|
|
|
|
|
|
|
} |
683
|
|
|
|
|
|
|
|
684
|
|
|
|
|
|
|
|
685
|
|
|
|
|
|
|
=head2 Prefix Analysis - B, B, B, B |
686
|
|
|
|
|
|
|
|
687
|
|
|
|
|
|
|
This method analyses text for the common prefixes of each active language. |
688
|
|
|
|
|
|
|
|
689
|
|
|
|
|
|
|
The methods are, respectively, for prefixes of size 1, 2, 3 and 4. |
690
|
|
|
|
|
|
|
|
691
|
|
|
|
|
|
|
=cut |
692
|
|
|
|
|
|
|
|
693
|
|
|
|
|
|
|
sub _langof_by_prefix_method { |
694
|
7
|
|
|
7
|
|
7103
|
use Text::Affixes; |
|
7
|
|
|
|
|
26870
|
|
|
7
|
|
|
|
|
1089
|
|
695
|
|
|
|
|
|
|
|
696
|
58
|
|
|
58
|
|
389
|
(my $method = shift) =~ /^prefixes(\d)$/; |
697
|
58
|
|
|
|
|
223
|
my $text = shift; |
698
|
|
|
|
|
|
|
|
699
|
58
|
|
|
|
|
607
|
my $prefixes = get_prefixes( {min => $1, max => $1}, $text); |
700
|
|
|
|
|
|
|
|
701
|
58
|
|
|
|
|
78680
|
return _langof_by_method($method, $$prefixes{$1}, $text); |
702
|
|
|
|
|
|
|
} |
703
|
|
|
|
|
|
|
|
704
|
|
|
|
|
|
|
=head2 Suffix Analysis - B, B, B, B |
705
|
|
|
|
|
|
|
|
706
|
|
|
|
|
|
|
Similar to the Prefix Analysis (see above), but instead analysing common |
707
|
|
|
|
|
|
|
suffixes. |
708
|
|
|
|
|
|
|
|
709
|
|
|
|
|
|
|
The methods are, respectively, for suffixes of size 1, 2, 3 and 4. |
710
|
|
|
|
|
|
|
|
711
|
|
|
|
|
|
|
=cut |
712
|
|
|
|
|
|
|
|
713
|
|
|
|
|
|
|
sub _langof_by_suffix_method { |
714
|
7
|
|
|
7
|
|
47
|
use Text::Affixes; |
|
7
|
|
|
|
|
14
|
|
|
7
|
|
|
|
|
886
|
|
715
|
|
|
|
|
|
|
|
716
|
61
|
|
|
61
|
|
389
|
(my $method = shift) =~ /^suffixes(\d)$/; |
717
|
61
|
|
|
|
|
293
|
my $text = shift; |
718
|
|
|
|
|
|
|
|
719
|
61
|
|
|
|
|
610
|
my $suffixes = get_suffixes({min => $1, max => $1}, $text); |
720
|
|
|
|
|
|
|
|
721
|
61
|
|
|
|
|
78743
|
return _langof_by_method($method, $$suffixes{$1}, $text); |
722
|
|
|
|
|
|
|
} |
723
|
|
|
|
|
|
|
|
724
|
|
|
|
|
|
|
### |
725
|
|
|
|
|
|
|
|
726
|
|
|
|
|
|
|
# Have you seen my brother? He's a two line long comment. I think he |
727
|
|
|
|
|
|
|
# might be lost... :-\ Me and my father have been looking for him for |
728
|
|
|
|
|
|
|
# some time now :-/ |
729
|
|
|
|
|
|
|
|
730
|
|
|
|
|
|
|
### |
731
|
|
|
|
|
|
|
|
732
|
|
|
|
|
|
|
=head2 Ngram Categorization - B, B, B, B |
733
|
|
|
|
|
|
|
|
734
|
|
|
|
|
|
|
Ngrams are sequences of tokens. You can think of them as syllables, but they |
735
|
|
|
|
|
|
|
are also more than that, as they are not only comprised by characters, but also |
736
|
|
|
|
|
|
|
by spaces (delimiting or separating words). |
737
|
|
|
|
|
|
|
|
738
|
|
|
|
|
|
|
Ngrams are a very good way for identifying languages, given that the most |
739
|
|
|
|
|
|
|
common ones of each language are not generally very common in others. |
740
|
|
|
|
|
|
|
|
741
|
|
|
|
|
|
|
This is usually the best method for small amounts of text or too many active |
742
|
|
|
|
|
|
|
languages. |
743
|
|
|
|
|
|
|
|
744
|
|
|
|
|
|
|
The methods are, respectively, for ngrams of size 1, 2, 3 and 4. |
745
|
|
|
|
|
|
|
|
746
|
|
|
|
|
|
|
=cut |
747
|
|
|
|
|
|
|
|
748
|
|
|
|
|
|
|
sub _langof_by_ngram_method { |
749
|
7
|
|
|
7
|
|
5687
|
use Text::Ngram qw(ngram_counts); |
|
7
|
|
|
|
|
22564
|
|
|
7
|
|
|
|
|
11309
|
|
750
|
|
|
|
|
|
|
|
751
|
58
|
|
|
58
|
|
339
|
(my $method = shift) =~ /^ngrams([1-4])$/; |
752
|
58
|
|
|
|
|
182
|
my $text = shift; |
753
|
|
|
|
|
|
|
|
754
|
58
|
|
|
|
|
445
|
my $ngrams = ngram_counts( {spaces => 0}, $text, $1); |
755
|
|
|
|
|
|
|
|
756
|
58
|
50
|
|
|
|
219808
|
$ngrams = 'letters' if $ngrams eq "ngrams1"; |
757
|
58
|
|
|
|
|
246
|
return _langof_by_method($method, $ngrams, $text); |
758
|
|
|
|
|
|
|
} |
759
|
|
|
|
|
|
|
|
760
|
|
|
|
|
|
|
=head1 LANGUAGE MANIPULATION |
761
|
|
|
|
|
|
|
|
762
|
|
|
|
|
|
|
When trying to perform language identification, C works not with |
763
|
|
|
|
|
|
|
all available languages, but instead with the ones that are active. |
764
|
|
|
|
|
|
|
|
765
|
|
|
|
|
|
|
By default, all available languages are active, but that can be changed by the |
766
|
|
|
|
|
|
|
user. |
767
|
|
|
|
|
|
|
|
768
|
|
|
|
|
|
|
For your convenience, several methods regarding language manipulation were |
769
|
|
|
|
|
|
|
created. In order to use them, load the module with the tag |
770
|
|
|
|
|
|
|
:language_manipulation. |
771
|
|
|
|
|
|
|
|
772
|
|
|
|
|
|
|
These methods work with the two letters code for languages. |
773
|
|
|
|
|
|
|
|
774
|
|
|
|
|
|
|
=over 6 |
775
|
|
|
|
|
|
|
|
776
|
|
|
|
|
|
|
=item B |
777
|
|
|
|
|
|
|
|
778
|
|
|
|
|
|
|
Activate a language |
779
|
|
|
|
|
|
|
|
780
|
|
|
|
|
|
|
activate_language('en'); |
781
|
|
|
|
|
|
|
|
782
|
|
|
|
|
|
|
# or |
783
|
|
|
|
|
|
|
|
784
|
|
|
|
|
|
|
activate_language($_) for get_all_languages(); |
785
|
|
|
|
|
|
|
|
786
|
|
|
|
|
|
|
=cut |
787
|
|
|
|
|
|
|
|
788
|
|
|
|
|
|
|
sub activate_language { |
789
|
1
|
50
|
|
1
|
1
|
11
|
unless (grep { $_ eq $_[0] } @active_languages) { |
|
0
|
|
|
|
|
0
|
|
790
|
1
|
|
|
|
|
4
|
push @active_languages, $_[0]; |
791
|
|
|
|
|
|
|
} |
792
|
1
|
|
|
|
|
5
|
return @active_languages; |
793
|
|
|
|
|
|
|
} |
794
|
|
|
|
|
|
|
|
795
|
|
|
|
|
|
|
=item B |
796
|
|
|
|
|
|
|
|
797
|
|
|
|
|
|
|
Activates all languages |
798
|
|
|
|
|
|
|
|
799
|
|
|
|
|
|
|
activate_all_languages(); |
800
|
|
|
|
|
|
|
|
801
|
|
|
|
|
|
|
=cut |
802
|
|
|
|
|
|
|
|
803
|
|
|
|
|
|
|
sub activate_all_languages { |
804
|
2
|
|
|
2
|
1
|
6
|
@active_languages = get_all_languages(); |
805
|
2
|
|
|
|
|
16
|
return @active_languages; |
806
|
|
|
|
|
|
|
} |
807
|
|
|
|
|
|
|
|
808
|
|
|
|
|
|
|
=item B |
809
|
|
|
|
|
|
|
|
810
|
|
|
|
|
|
|
Deactivates a language |
811
|
|
|
|
|
|
|
|
812
|
|
|
|
|
|
|
deactivate_language('en'); |
813
|
|
|
|
|
|
|
|
814
|
|
|
|
|
|
|
=cut |
815
|
|
|
|
|
|
|
|
816
|
|
|
|
|
|
|
sub deactivate_language { |
817
|
1
|
|
|
1
|
1
|
5
|
@active_languages = grep { ! ($_ eq $_[0]) } @active_languages; |
|
27
|
|
|
|
|
40
|
|
818
|
1
|
|
|
|
|
13
|
return @active_languages; |
819
|
|
|
|
|
|
|
} |
820
|
|
|
|
|
|
|
|
821
|
|
|
|
|
|
|
=item B |
822
|
|
|
|
|
|
|
|
823
|
|
|
|
|
|
|
Deactivates all languages |
824
|
|
|
|
|
|
|
|
825
|
|
|
|
|
|
|
deactivate_all_languages(); |
826
|
|
|
|
|
|
|
|
827
|
|
|
|
|
|
|
=cut |
828
|
|
|
|
|
|
|
|
829
|
|
|
|
|
|
|
sub deactivate_all_languages { |
830
|
3
|
|
|
3
|
1
|
15
|
@active_languages = (); |
831
|
3
|
|
|
|
|
13
|
return @active_languages; |
832
|
|
|
|
|
|
|
} |
833
|
|
|
|
|
|
|
|
834
|
|
|
|
|
|
|
=item B |
835
|
|
|
|
|
|
|
|
836
|
|
|
|
|
|
|
Returns the names of all available languages |
837
|
|
|
|
|
|
|
|
838
|
|
|
|
|
|
|
my @all_languages = get_all_languages(); |
839
|
|
|
|
|
|
|
|
840
|
|
|
|
|
|
|
=cut |
841
|
|
|
|
|
|
|
|
842
|
|
|
|
|
|
|
sub get_all_languages { |
843
|
55
|
|
|
55
|
1
|
1067
|
return @all_languages; |
844
|
|
|
|
|
|
|
} |
845
|
|
|
|
|
|
|
|
846
|
|
|
|
|
|
|
=item B |
847
|
|
|
|
|
|
|
|
848
|
|
|
|
|
|
|
Returns the names of all active languages |
849
|
|
|
|
|
|
|
|
850
|
|
|
|
|
|
|
my @active_languages = get_active_languages(); |
851
|
|
|
|
|
|
|
|
852
|
|
|
|
|
|
|
=cut |
853
|
|
|
|
|
|
|
|
854
|
|
|
|
|
|
|
sub get_active_languages { |
855
|
367
|
|
|
367
|
1
|
2873
|
return @active_languages; |
856
|
|
|
|
|
|
|
} |
857
|
|
|
|
|
|
|
|
858
|
|
|
|
|
|
|
=item B |
859
|
|
|
|
|
|
|
|
860
|
|
|
|
|
|
|
Returns the names of all inactive languages |
861
|
|
|
|
|
|
|
|
862
|
|
|
|
|
|
|
my @active_languages = get_inactive_languages(); |
863
|
|
|
|
|
|
|
|
864
|
|
|
|
|
|
|
=cut |
865
|
|
|
|
|
|
|
|
866
|
|
|
|
|
|
|
sub get_inactive_languages { |
867
|
2
|
|
|
2
|
1
|
739
|
return grep { ! is_active($_) } get_all_languages(); |
|
54
|
|
|
|
|
67
|
|
868
|
|
|
|
|
|
|
} |
869
|
|
|
|
|
|
|
|
870
|
|
|
|
|
|
|
=item B |
871
|
|
|
|
|
|
|
|
872
|
|
|
|
|
|
|
Returns the name of the language if it is active, an empty list otherwise |
873
|
|
|
|
|
|
|
|
874
|
|
|
|
|
|
|
if (is_active('en')) { |
875
|
|
|
|
|
|
|
# YOUR CODE HERE |
876
|
|
|
|
|
|
|
} |
877
|
|
|
|
|
|
|
|
878
|
|
|
|
|
|
|
=cut |
879
|
|
|
|
|
|
|
|
880
|
|
|
|
|
|
|
sub is_active { |
881
|
56
|
|
|
56
|
1
|
73
|
return grep { $_ eq $_[0] } get_active_languages(); |
|
729
|
|
|
|
|
854
|
|
882
|
|
|
|
|
|
|
} |
883
|
|
|
|
|
|
|
|
884
|
|
|
|
|
|
|
=item B |
885
|
|
|
|
|
|
|
|
886
|
|
|
|
|
|
|
Returns the name of the language if it exists, an empty list otherwise |
887
|
|
|
|
|
|
|
|
888
|
|
|
|
|
|
|
if (is_valid_language('en')) { |
889
|
|
|
|
|
|
|
# YOUR CODE HERE |
890
|
|
|
|
|
|
|
} |
891
|
|
|
|
|
|
|
|
892
|
|
|
|
|
|
|
=cut |
893
|
|
|
|
|
|
|
|
894
|
|
|
|
|
|
|
sub is_valid_language { |
895
|
37
|
|
|
37
|
1
|
13243
|
return grep { $_ eq $_[0] } get_all_languages(); |
|
999
|
|
|
|
|
1359
|
|
896
|
|
|
|
|
|
|
} |
897
|
|
|
|
|
|
|
|
898
|
|
|
|
|
|
|
=item B |
899
|
|
|
|
|
|
|
|
900
|
|
|
|
|
|
|
Sets the active languages |
901
|
|
|
|
|
|
|
|
902
|
|
|
|
|
|
|
set_active_languages('en', 'pt'); |
903
|
|
|
|
|
|
|
|
904
|
|
|
|
|
|
|
# or |
905
|
|
|
|
|
|
|
|
906
|
|
|
|
|
|
|
set_active_languages(get_all_languages()); |
907
|
|
|
|
|
|
|
|
908
|
|
|
|
|
|
|
=cut |
909
|
|
|
|
|
|
|
|
910
|
|
|
|
|
|
|
sub set_active_languages { |
911
|
3
|
|
|
3
|
1
|
11
|
@active_languages = grep { is_valid_language($_) } @_; |
|
6
|
|
|
|
|
18
|
|
912
|
3
|
|
|
|
|
25
|
return @active_languages; |
913
|
|
|
|
|
|
|
} |
914
|
|
|
|
|
|
|
|
915
|
|
|
|
|
|
|
=item B |
916
|
|
|
|
|
|
|
|
917
|
|
|
|
|
|
|
Given the two letter tag of a language, returns its name |
918
|
|
|
|
|
|
|
|
919
|
|
|
|
|
|
|
my $language_name = name_of('pt'); |
920
|
|
|
|
|
|
|
|
921
|
|
|
|
|
|
|
=cut |
922
|
|
|
|
|
|
|
|
923
|
|
|
|
|
|
|
sub name_of { |
924
|
1
|
|
50
|
1
|
1
|
4
|
my $tag = shift || return undef; |
925
|
1
|
|
|
|
|
7
|
return $languages{_names}{$tag}; |
926
|
|
|
|
|
|
|
} |
927
|
|
|
|
|
|
|
|
928
|
|
|
|
|
|
|
=back |
929
|
|
|
|
|
|
|
|
930
|
|
|
|
|
|
|
=cut |
931
|
|
|
|
|
|
|
|
932
|
|
|
|
|
|
|
1; |
933
|
|
|
|
|
|
|
__END__ |